[go: up one dir, main page]

WO2018139275A1 - Appareil, dispositif de sortie, procédé de sortie et programme de sortie - Google Patents

Appareil, dispositif de sortie, procédé de sortie et programme de sortie Download PDF

Info

Publication number
WO2018139275A1
WO2018139275A1 PCT/JP2018/001082 JP2018001082W WO2018139275A1 WO 2018139275 A1 WO2018139275 A1 WO 2018139275A1 JP 2018001082 W JP2018001082 W JP 2018001082W WO 2018139275 A1 WO2018139275 A1 WO 2018139275A1
Authority
WO
WIPO (PCT)
Prior art keywords
action
unit
command
utterance
correction
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/JP2018/001082
Other languages
English (en)
Japanese (ja)
Inventor
薫 木下
清水　雅芳
真司 神田
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Fujitsu Ltd
Original Assignee
Fujitsu Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Fujitsu Ltd filed Critical Fujitsu Ltd
Publication of WO2018139275A1 publication Critical patent/WO2018139275A1/fr
Priority to US16/298,158 priority Critical patent/US10916236B2/en
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L15/00—Speech recognition
    • G10L15/02—Feature extraction for speech recognition; Selection of recognition unit
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J11/00—Manipulators not otherwise provided for
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J11/00—Manipulators not otherwise provided for
    • B25J11/0005—Manipulators having means for high-level communication with users, e.g. speech generator, face recognition means
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/279—Recognition of textual entities
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/30—Semantic analysis
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L15/00—Speech recognition
    • G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue

Definitions

  • This case relates to equipment, output device, output method and output program.
  • a robot that expresses a combination of speech and action has been developed.
  • a technique is disclosed in which the same action is not performed using a random number in order to avoid selecting a uniform action that does not change for the same phrase (for example, , See Patent Document 1).
  • a technique for performing an operation according to a history of input information in an animal type robot is disclosed (for example, see Patent Document 2).
  • a technique using random numbers does not guarantee a natural action. For example, if the action “Looks like” is repeated, if the nod is slightly strong in the first time, and if the nod is slightly loosed in the second time, the action may not be uniform but unnatural.
  • an action is determined based on the accumulation of past information, and thus there is a possibility that an unnatural action may occur in relation to the utterance after the action.
  • the object of the present invention is to provide a device, an output device, an output method, and an output program in which an action corresponding to the utterance content is realized.
  • the output device includes an acquisition unit that acquires an utterance command and an action command, an analysis unit that analyzes an utterance content after an action based on the action command, and an analysis of the analysis unit.
  • a correction unit that corrects the action command according to a result; an output unit that outputs the action command corrected by the correction unit; and the utterance command.
  • the device analyzes an utterance content after an action based on the action command with respect to the utterance unit that utters according to the utterance command, an action unit that performs an action according to the action command, and the utterance command.
  • (A) is a block diagram for demonstrating the apparatus with which the apparatus control apparatus which concerns on Example 1 is applied
  • (b) is a figure which illustrates the hardware constitutions of an apparatus control apparatus.
  • (A) is a block diagram of functions realized by executing the device control program
  • (b) is a block diagram of an action correction unit
  • (c) is a diagram illustrating a table stored in an action database. is there.
  • (A) is a figure which illustrates each function implement
  • (b) is a block diagram of an action correction part
  • (c) illustrates the calculated parameter reflected value It is a figure to do.
  • (A) is a figure which illustrates the table stored in the keyword database
  • (b) is a figure which illustrates extraction of the parameter which concerns on emotion.
  • (A) is a figure which illustrates each function implement
  • (b) is a figure which illustrates the table stored in the action database.
  • FIG. 1A is a block diagram for explaining a device 300 to which the device control apparatus 100 according to the first embodiment is applied.
  • the device 300 includes a device control apparatus 100 and a robot 200.
  • the robot 200 is a speech robot and includes a voice output unit 201, an action unit 202, and the like.
  • the audio output unit 201 is a speaker or the like and speaks according to audio data sent from the device control apparatus 100.
  • the action unit 202 includes a motor that drives a joint of the robot 200 and performs an action by realizing a joint angle according to motor control data sent from the device control apparatus 100.
  • the user of the device 300 is, for example, a developer who develops a demonstration that combines speech and action using the robot 200.
  • FIG. 1B is a diagram illustrating a hardware configuration of the device control apparatus 100.
  • the device control apparatus 100 includes a CPU 101, a RAM 102, a storage device 103, an input device 104, an interface 105, and the like. Each of these devices is connected by a bus or the like.
  • a CPU (Central Processing Unit) 101 is a central processing unit.
  • the CPU 101 includes one or more cores.
  • a RAM (Random Access Memory) 102 is a volatile memory that temporarily stores programs executed by the CPU 101, data processed by the CPU 101, and the like.
  • the storage device 103 is a nonvolatile storage device. As the storage device 103, for example, a ROM (Read Only Memory), a solid state drive (SSD) such as a flash memory, a hard disk driven by a hard disk drive, or the like can be used.
  • the input device 104 is a keyboard, a mouse, or the like, and is a device for a user to input information.
  • the interface 105 is an interface for transmitting and receiving signals to and from the robot 200.
  • FIG. 2A is a block diagram of functions realized by executing the device control program. Each function may be hardware such as a dedicated circuit.
  • a script acquisition unit 10 a script analysis unit 20, a voice data output unit 30, an utterance analysis unit 40, an action identification unit 50, an action correction unit 60, an action A data output unit 70, a motor control unit 80, and the like are realized.
  • the action correction unit 60 includes an action selection unit 61 and an action database 62.
  • the script acquisition unit 10 acquires, for example, a script input from the input device 104 by the user.
  • the script includes an utterance command and an action command.
  • the utterance command includes the first utterance content and the second utterance content.
  • “Today is tired” is input as the first utterance content
  • “Tomorrow is also very likely” is input as the second utterance content after the first utterance content.
  • “gakkuri” is input as an action command between the first utterance content and the second utterance content.
  • the script acquisition unit 10 sends the acquired script to the script analysis unit 20.
  • the script analysis unit 20 analyzes the script received from the script acquisition unit 10.
  • the script analysis unit 20 extracts the first utterance content and the second utterance content from the script received from the script acquisition unit 10 and sends them to the voice data output unit 30.
  • the voice data output unit 30 sends the voice data of the first utterance content and the second utterance content to the voice output unit 201 of FIG.
  • the robot 200 utters “Today is tired” from the voice output unit 201, and then utters “Tomorrow is very promising”.
  • the script analysis unit 20 sends the first utterance content and the second utterance content to the utterance analysis unit 40. Further, the script analysis unit 20 extracts an action command from the script received from the script acquisition unit 10 and sends it to the action identification unit 50.
  • the utterance analysis unit 40 analyzes the first utterance content and the second utterance content. For example, the utterance analysis unit 40 refers to the keyword database 41 to extract the first keyword from the first utterance content. For example, “tired” is extracted as the first keyword.
  • the utterance analysis unit 40 sends the first keyword or information indicating that there is no keyword in the first utterance content to the action correction unit 60 as an index A.
  • the utterance analysis unit 40 refers to the keyword database 41 to extract the second keyword from the second utterance content. For example, “Tomorrow is also” is extracted as the second keyword.
  • the utterance analysis unit 40 sends the second keyword or information indicating that there is no keyword in the second utterance content to the action correction unit 60 as an index B. Index A and index B have a function as a correction method.
  • the action identification unit 50 refers to the action database 51 to identify the basic operation of the action command “gakkuri” received from the script analysis unit 20.
  • the action specifying unit 50 specifies the nodding operation as the basic operation of the action command “gakkuri”.
  • the action identification unit 50 identifies an action correction method for a nodding operation. For example, for a nodding motion, the angles of the neck, hand, arm, etc. are specified as the correction method.
  • the action identification unit 50 sends the identified action command and action modification method to the action modification unit 60.
  • the action database 62 stores correction examples corresponding to the index A and the index B as correction contents for each of a plurality of actions. For example, when the action command is “I like it”, the angles of the neck, hand, arm, etc. corresponding to the index A and the index B are stored as a correction example. For example, if “Fatigue” is included in the index A and “Tomorrow” is also included in the index B, it is conceivable to increase the motion (joint angle) of nodding.
  • the action selection unit 61 corrects the basic operation of the action command “gakkuri” by selecting an action example corresponding to the index A and the index B from the action database 62.
  • the action selection unit 61 sends the corrected operation to the action data output unit 70.
  • the action data output unit 70 sends data related to the corrected operation to the motor control unit 80.
  • the motor control unit 80 controls the motor provided in the action unit 202 in FIG. 1A based on the received data, thereby causing the robot 200 to perform an action. Thereby, after the voice output unit 201 utters “I'm tired today”, the robot 200 performs the action “correctly” in which the action unit 202 is corrected. “It ’s very promising.”
  • the action command is modified according to the utterance content after the action based on the action command.
  • the action unit 202 of the robot 200 performs an action in response to the corrected action command. If only the utterance content before the action is analyzed, the action may become unnatural depending on the relationship with the utterance content after the action. On the other hand, in the present embodiment, since the analysis result of the utterance content after the action is reflected in the correction of the action command, an appropriate action corresponding to the utterance content is performed. In addition, the action can be corrected more appropriately by correcting the action command according to the utterance content before and after the action.
  • the voice data output unit 30 sends voice data to the voice output unit 201 to cause the robot 200 to speak
  • the action data output unit 70 sends data related to the corrected operation to the motor control unit 80. Sending the robot 200 to perform an action is not limited to this.
  • the audio data output from the audio data output unit 30 and the data related to the operation output from the action data output unit 70 may be stored in a storage device such as a memory.
  • the script acquisition unit 10 functions as an example of an acquisition unit that acquires an utterance command and an action command.
  • the utterance analysis unit 40 functions as an example of an analysis unit that analyzes the utterance content after the action based on the action command with respect to the utterance command.
  • the action correction unit 60 functions as an example of a correction unit that corrects the action command according to the analysis result of the analysis unit.
  • the voice data output unit 30 and the action data output unit 70 function as an example of an output unit that outputs the action command modified by the modification unit and the utterance command.
  • the voice output unit 201 functions as an example of an utterance unit that utters in response to an utterance command.
  • the action unit 202 functions as an example of an action unit that performs an action according to an action command.
  • FIG. 3A is a diagram illustrating functions implemented in the device control apparatus 100a according to the second embodiment.
  • the difference between the device control device 100a and the device control device 100 of the first embodiment is that an utterance analysis unit 40a is realized instead of the utterance analysis unit 40, an action specification unit 50a is realized instead of the action specification unit 50, and an action correction An action correcting unit 60a is realized instead of the unit 60.
  • the action correction unit 60 a includes an action database 62, a reflection value calculation unit 63, and a correction unit 64.
  • the utterance analysis unit 40 a refers to the keyword database 41 to extract the first keyword from the first utterance content. For example, “tired” is extracted as the first keyword.
  • the utterance analysis unit 40a refers to the keyword database 41 to extract the second keyword from the second utterance content. For example, “Tomorrow is also” is extracted as the second keyword.
  • the action identification unit 50a refers to the action database 51 to identify an emotion corresponding to the action command “gakkuri” received from the script analysis unit 20, and sends it to the utterance analysis unit 40a. For example, it is assumed that “sorrow” is associated with “sorrow” among emotions.
  • the utterance analysis unit 40a calculates the correction amount as a parameter based on the first keyword, the second keyword, and the emotion “sorrow” corresponding to the action command “gakkuri”.
  • FIG. 4A is a diagram illustrating a table stored in the keyword database 41.
  • each keyword is associated with a parameter related to one or more emotions such as emotions.
  • “joy” is associated with the parameter value “15” of “joy” and the parameter value “35” of “joy”.
  • “fatigue” is associated with a parameter value “15” of “anger” and a parameter value “35” of “sorrow”.
  • a coefficient “ ⁇ 1” for strengthening emotion is associated as a parameter.
  • a coefficient “ ⁇ ⁇ 1” for relieving emotion is associated as a parameter.
  • the speech analysis unit 40a extracts the parameter value “15” for “anger” and the parameter value “35” for “sorrow” for “fatigue”. Further, the speech analysis unit 40a extracts the coefficient “ ⁇ 1” for “tomorrow”, multiplies each parameter value by “ ⁇ 1” as a coefficient, and sets the parameter values “15” and “sorrow” for “anger”. The parameter value “35” is extracted. The utterance analysis unit 40a extracts and adds the parameter values “35” and “35” of the emotion “sorrow” received from the action specifying unit 50a. The speech analysis unit 40a obtains the obtained numerical value “70” as a correction amount.
  • the parameter value “16” for “joy” is extracted for “fun”, and the parameter value “36” for “enjoy” is extracted.
  • the action specifying unit 50a specifies the emotion “Raku”
  • the parameter values “0” and “36” of “Raku” are added together, and the obtained numerical value “36” is extracted as the correction amount.
  • one keyword is extracted from each of the first utterance content and the second utterance content, but the present invention is not limited to this.
  • the first utterance content includes a plurality of keywords
  • the parameter values of the keywords may be extracted and the corresponding parameters may be added together.
  • the action specifying unit 50a specifies the basic operation of the action command “gakkuri” received from the script analysis unit 20 by referring to the action database 51.
  • the action specifying unit 50a specifies the joint angles of the neck, hand, arm, etc., for realizing the basic action of “gakkuri”.
  • the action specifying unit 50 a refers to the action database 51 to acquire the correction reflection degree of each joint with respect to the action command “gakkuri”.
  • the action specifying unit 50a sends the specified basic operation and the correction reflection degree to the action correcting unit 60a.
  • the correction reflection degree has a function as a correction method.
  • the reflection value calculation unit 63 calculates a parameter reflection value from the basic operation and the correction reflection degree received from the action specifying unit 50a and the correction value received from the utterance analysis unit 40a.
  • FIG. 3C is a diagram illustrating the calculated parameter reflection value. Since the correction reflection degree for the joint 1 is “0”, the operation amount is not corrected regardless of the correction amount. For the joint 2, since the correction reflection degree is “1”, “1” is multiplied by the correction amount 70, the obtained correction value “70” is added to the operation amount “40”, and the correction operation amount “110”. Is calculated. For the joint 3, since the correction reflection degree is “2”, the correction amount 70 is multiplied by “2”, the obtained correction value “140” is added to the operation amount “10”, and the correction operation amount “150”.
  • the correction unit 64 sends the corrected operation to the action data output unit 70.
  • the action command is modified according to the utterance content after the action based on the action command.
  • the action unit 202 of the robot 200 performs an action in response to the corrected action command. If only the utterance content before the action is analyzed, the action may become unnatural depending on the relationship with the utterance content after the action.
  • the analysis result of the utterance content after the action is reflected in the correction of the action command, an appropriate action corresponding to the utterance content is performed.
  • the action can be corrected more appropriately by correcting the action command according to the utterance content before and after the action.
  • the parameter related to emotion is used as the correction amount of the basic motion, the action can be corrected according to the emotion.
  • the voice data output unit 30 sends voice data to the voice output unit 201 to cause the robot 200 to speak
  • the action data output unit 70 sends data related to the corrected operation to the motor control unit 80. Sending the robot 200 to perform an action is not limited to this.
  • the audio data output from the audio data output unit 30 and the data related to the operation output from the action data output unit 70 may be stored in a storage device such as a memory.
  • the script acquisition unit 10 functions as an example of an acquisition unit that acquires an utterance command and an action command.
  • the utterance analysis unit 40a functions as an example of an analysis unit that analyzes the utterance content after the action based on the action command with respect to the utterance command.
  • the action correction unit 60a functions as an example of a correction unit that corrects the action command according to the analysis result of the analysis unit.
  • the voice data output unit 30 and the action data output unit 70 function as an example of an output unit that outputs the action command modified by the modification unit and the utterance command.
  • the voice output unit 201 functions as an example of an utterance unit that utters in response to an utterance command.
  • the action unit 202 functions as an example of an action unit that performs an action according to an action command.
  • FIG. 5A is a diagram illustrating functions implemented in the device control apparatus 100b according to the third embodiment.
  • the difference between the device control device 100b and the device control device 100a of the second embodiment is that an utterance analysis unit 40b is provided instead of the utterance analysis unit 40a, an action specifying unit 50b is provided instead of the action specifying unit 50a, and a new keyword analysis is performed.
  • the condition creation unit 43 and the action database 44 are provided. Also in this example, “Today is tired” is input as the first utterance content, “Tomorrow is also very promising” is input as the second utterance content, and “Gakkaku” is input as the action command. The case will be described. About the function similar to Example 2, description is abbreviate
  • the action identification unit 50b identifies the basic operation of the action command “gakkuri” received from the script analysis unit 20 by referring to the action database 51.
  • the action specifying unit 50b specifies joint angles such as a neck, a hand, and an arm for realizing a basic action of “gakkuri”.
  • the action specifying unit 50 b refers to the action database 51 to acquire the correction reflection degree for each joint with respect to the action command “gakkuri”.
  • the action specifying unit 50b sends the specified basic operation and correction reflection degree to the action correcting unit 60a.
  • the action specifying unit 50 b sends the specified action command to the keyword analysis condition creating unit 43.
  • FIG. 5B is a diagram illustrating a table stored in the action database 44.
  • the table stores the relevance levels of emotions in association with each action command.
  • the degree of relevance indicates that the higher the numerical value, the higher the relevance to each emotion.
  • special information for prioritizing conditions other than the basic parameters such as emotions in association with each action command may be stored as other parameters.
  • the keyword analysis condition creating unit 43 refers to the action database 44, and acquires a combination of relevance levels of emotions corresponding to the action command received from the action specifying unit 50a, or special information.
  • the keyword analysis condition creation unit 43 sends a combination of the relevance levels of emotions or special information or special information to the utterance analysis unit 40b.
  • the utterance analysis unit 40 b refers to the keyword database 41 to extract the first keyword from the first utterance content. For example, “tired” is extracted as the first keyword.
  • the utterance analysis unit 40 b refers to the keyword database 41 to extract the second keyword from the second utterance content. For example, “Tomorrow is also” is extracted as the second keyword.
  • the utterance analysis unit 40b extracts parameters for items whose relevance degree is equal to or greater than a threshold among the combinations of emotions received from the keyword analysis condition creation unit 43. For example, in the action command “gakkuri”, it is assumed that the degree of relevance of “sorrow” is set to a threshold value or more. In the example described with reference to FIG. 4A, the utterance analysis unit 40b extracts the parameter value “35” of “sorrow” for “tired”. Further, the utterance analysis unit 40b extracts the parameter value “35” of “sorrow” by multiplying each parameter value by “ ⁇ 1” as a coefficient for “tomorrow”. The utterance analysis unit 40b extracts the parameter values “35” and “35” of these “sorrows” and adds them. The speech analysis unit 40b obtains the obtained numerical value “70” as a correction amount.
  • the action command is modified according to the utterance content after the action based on the action command.
  • the action unit 202 of the robot 200 performs an action in response to the corrected action command. If only the utterance content before the action is analyzed, the action may become unnatural depending on the relationship with the utterance content after the action.
  • the analysis result of the utterance content after the action is reflected in the correction of the action command, an appropriate action corresponding to the utterance content is performed.
  • the action can be corrected more appropriately by correcting the action command according to the utterance content before and after the action.
  • the parameter related to emotion is used as the correction amount of the basic motion, the action can be corrected according to the emotion.
  • the voice data output unit 30 sends voice data to the voice output unit 201 to cause the robot 200 to speak
  • the action data output unit 70 sends data related to the corrected operation to the motor control unit 80. Sending the robot 200 to perform an action is not limited to this.
  • the audio data output from the audio data output unit 30 and the data related to the operation output from the action data output unit 70 may be stored in a storage device such as a memory.
  • the script acquisition unit 10 functions as an example of an acquisition unit that acquires an utterance command and an action command.
  • the utterance analysis unit 40b functions as an example of an analysis unit that analyzes the utterance content after the action based on the action command with respect to the utterance command.
  • the action correction unit 60b functions as an example of a correction unit that corrects the action command according to the analysis result of the analysis unit.
  • the voice data output unit 30 and the action data output unit 70 function as an example of an output unit that outputs the action command modified by the modification unit and the utterance command.
  • the voice output unit 201 functions as an example of an utterance unit that utters in response to an utterance command.
  • the action unit 202 functions as an example of an action unit that performs an action according to an action command.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Computational Linguistics (AREA)
  • Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Human Computer Interaction (AREA)
  • Robotics (AREA)
  • Mechanical Engineering (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Acoustics & Sound (AREA)
  • Multimedia (AREA)
  • Manipulator (AREA)

Abstract

L'invention porte sur un appareil, un dispositif de sortie, un procédé de sortie et un programme de sortie qui permettent de réaliser une action correspondant au contenu vocal. Le dispositif de sortie comporte une unité d'acquisition (10) qui acquiert une commande vocale et une commande d'action; une unité d'analyse (40) qui, par rapport à la commande vocale, analyse le contenu vocal qui suit l'action sur la base de la commande d'action; une unité de correction (60) qui corrige la commande d'action en fonction des résultats d'analyse de l'unité d'analyse (40); et une unité de sortie (70) qui délivre en sortie la commande vocale et la commande d'action corrigée par l'unité de correction (60).
PCT/JP2018/001082 2017-01-30 2018-01-17 Appareil, dispositif de sortie, procédé de sortie et programme de sortie Ceased WO2018139275A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US16/298,158 US10916236B2 (en) 2017-01-30 2019-03-11 Output device, output method, and electronic apparatus

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2017-014042 2017-01-30
JP2017014042A JP6729424B2 (ja) 2017-01-30 2017-01-30 機器、出力装置、出力方法および出力プログラム

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US16/298,158 Continuation US10916236B2 (en) 2017-01-30 2019-03-11 Output device, output method, and electronic apparatus

Publications (1)

Publication Number Publication Date
WO2018139275A1 true WO2018139275A1 (fr) 2018-08-02

Family

ID=62979278

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2018/001082 Ceased WO2018139275A1 (fr) 2017-01-30 2018-01-17 Appareil, dispositif de sortie, procédé de sortie et programme de sortie

Country Status (3)

Country Link
US (1) US10916236B2 (fr)
JP (1) JP6729424B2 (fr)
WO (1) WO2018139275A1 (fr)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10621983B2 (en) * 2018-04-20 2020-04-14 Spotify Ab Systems and methods for enhancing responsiveness to utterances having detectable emotion
EP4198978A1 (fr) * 2021-12-16 2023-06-21 Bull SAS Procédé, dispositif et programme informatique pour la reconnaissance d'émotions à partir d'un signal audio en temps réel

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2004268235A (ja) * 2003-03-11 2004-09-30 Sony Corp ロボット装置、その行動制御方法及びプログラム
JP2007050461A (ja) * 2005-08-17 2007-03-01 Nec Corp ロボット制御システム、ロボット装置、およびロボット制御方法
JP2010188429A (ja) * 2009-02-16 2010-09-02 Advanced Telecommunication Research Institute International 道案内ロボット
JP2014504959A (ja) * 2011-01-26 2014-02-27 本田技研工業株式会社 人間型ロボットのジェスチャと音声の同期
JP2015138147A (ja) * 2014-01-22 2015-07-30 シャープ株式会社 サーバ、対話装置、対話システム、対話方法および対話プログラム
JP2016126294A (ja) * 2015-01-08 2016-07-11 シャープ株式会社 音声対話制御装置、音声対話制御装置の制御方法、および音声対話装置

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6337552B1 (en) 1999-01-20 2002-01-08 Sony Corporation Robot apparatus
JP4465768B2 (ja) * 1999-12-28 2010-05-19 ソニー株式会社 音声合成装置および方法、並びに記録媒体
JP2002304188A (ja) * 2001-04-05 2002-10-18 Sony Corp 単語列出力装置および単語列出力方法、並びにプログラムおよび記録媒体
JP6122792B2 (ja) * 2014-02-06 2017-04-26 日本電信電話株式会社 ロボット制御装置、ロボット制御方法及びロボット制御プログラム
WO2017200074A1 (fr) * 2016-05-20 2017-11-23 日本電信電話株式会社 Procédé de dialogue, système de dialogue, dispositif de dialogue et programme
CN107030691B (zh) * 2017-03-24 2020-04-14 华为技术有限公司 一种看护机器人的数据处理方法及装置
US11107472B2 (en) * 2017-03-31 2021-08-31 Intel Corporation Management of human-machine dialogue involving multiple parties

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2004268235A (ja) * 2003-03-11 2004-09-30 Sony Corp ロボット装置、その行動制御方法及びプログラム
JP2007050461A (ja) * 2005-08-17 2007-03-01 Nec Corp ロボット制御システム、ロボット装置、およびロボット制御方法
JP2010188429A (ja) * 2009-02-16 2010-09-02 Advanced Telecommunication Research Institute International 道案内ロボット
JP2014504959A (ja) * 2011-01-26 2014-02-27 本田技研工業株式会社 人間型ロボットのジェスチャと音声の同期
JP2015138147A (ja) * 2014-01-22 2015-07-30 シャープ株式会社 サーバ、対話装置、対話システム、対話方法および対話プログラム
JP2016126294A (ja) * 2015-01-08 2016-07-11 シャープ株式会社 音声対話制御装置、音声対話制御装置の制御方法、および音声対話装置

Also Published As

Publication number Publication date
US10916236B2 (en) 2021-02-09
JP2018122366A (ja) 2018-08-09
US20190206387A1 (en) 2019-07-04
JP6729424B2 (ja) 2020-07-22

Similar Documents

Publication Publication Date Title
EP3192070B1 (fr) Conversion de texte en parole avec contenu émotionnel
US20200233908A1 (en) Interactive system and computer program therefor
CN110570879A (zh) 基于情绪识别的智能会话方法、装置及计算机设备
JP7462739B2 (ja) シーケンス－シーケンス・ニューラル・モデルにおける構造保持注意機構
US9099091B2 (en) Method and apparatus of adaptive textual prediction of voice data
EP1091303A3 (fr) Procédé et système pour fournir des alternatives aux textes derivés de sources stochastiques
KR101131278B1 (ko) 대화 로그를 이용한 학습 기반 대화 시스템 성능 향상 방법 및 그 장치
US20200074996A1 (en) Speech recognition apparatus, speech recognition method, and a recording medium
KR102153220B1 (ko) 동일성 판단에 기초한 음성 인식 결과 출력 방법 및 이를 이용한 장치
CN111613219B (zh) 语音数据识别方法、设备及介质
CN112331229A (zh) 语音检测方法、装置、介质和计算设备
CN114595314A (zh) 融合情绪的对话应答方法、装置、终端及存储装置
CN110619888A (zh) 一种ai语音速率调整方法、装置及电子设备
KR102170978B1 (ko) 음성 명령 및 증강 현실을 이용한 코딩 교육 장치
WO2018139275A1 (fr) Appareil, dispositif de sortie, procédé de sortie et programme de sortie
KR102758478B1 (ko) 음성 인식 방법 및 장치
JP6082657B2 (ja) ポーズ付与モデル選択装置とポーズ付与装置とそれらの方法とプログラム
US20240320519A1 (en) Systems and methods for providing a digital human in a virtual environment
KR102220106B1 (ko) 음성 인식된 문장의 보정 방법
KR20240014257A (ko) 합성 음성을 생성하는 방법 및 음성 합성 시스템
JP7291099B2 (ja) 音声認識方法及び装置
KR20220071523A (ko) 문자들의 시퀀스를 분할하는 방법 및 음성 합성 시스템
WO2025020166A1 (fr) Système et procédé permettant de transformer un audio à l'aide d'un modèle d'apprentissage automatique génératif de fondation
KR102711295B1 (ko) 음성 인식 장치 및 그것을 이용한 음성 인식 속도 개선 방법
KR20240014250A (ko) 스펙트로그램에 대응하는 어텐션 얼라인먼트의 인코더 스코어를 연산하는 방법 및 음성 합성 시스템

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18744432

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18744432

Country of ref document: EP

Kind code of ref document: A1